Publication

Abstract : Face recognition is a widely used authentication technique for various uses like Attendance marking, Human computer interaction, Security access, E-learning[23], online proctoring [22] [25] etc. Here we use face recognition for marking attendance for large groups of people together in a classroom[14] [24]. Compared to other traditional attendance marking methods face recognition will provide better accuracy and prevents proxy attendance, but it is a complex and time consuming process when it applied to a group of people simultaneously. In this proposed system we are focusing on reducing the complexity of the group face recognition process by finding a suitable scenario which will improve the accuracy and reduce the time complexity. We have used the advanced deep learning convolutional neural network models for face recognition. Neural networks will convert the facial features to vector embeddings and uses for face recognition. System has a final classification layer which will recognize the person and mark their presence in the attendance marking system. As per the analysis we have noticed that by marking attendance a group of 10 to 12 faces one can achieve maximum accuracy in less processing time.